The Washington Post reports on a study which contends that “most diversity training efforts at American companies are ineffective and even counterproductive.”

The article estimates that U.S. businesses spend $200 million to $300 million a year on diversity training. I can’t help thinking that the same tough analysis goes into justifying this effect as goes into, say, figuring out bonuses for CEOs.

Alexandra Kalev of the U. of Arizona led a review of 31 years’ worth of data from over 800 companies. Says the Post article:

…The analysis did not find that all diversity training is useless. Rather, it showed that mandatory programs — often undertaken mainly with an eye to avoiding liability in discrimination lawsuits — were the problem. When diversity training is voluntary and undertaken to advance a company’s business goals, it was associated with increased diversity in management.

Or, as Daniel Patrick Moynihan once remarked, “Everyone is entitled to their own opinion, but not their own facts.”

Kalev and her colleagues found that the best programs focused not on attitudes but on organizational skills — like establishing mentoring relationships, or giving women and minorities to take on high-profile roles.

I found an earlier study by Kalev and colleagues, Cracking the Glass Cages? Work Teams, Cross-Training and Management Diversity (this link is to a draft; the final version appeared in the American Journal of Sociology). One conclusion is that movement away from traditional hierarchical job structures — e.g., to self-directed work teams and to job-rotation programs — leads to increased gender and racial diversity in management.